使用离体循环人全血测定对双特异性抗体进行临床前免疫毒性评估

The technologies for producing bispecific antibodies (BsAbs) are relatively new, and only recently have several BsAbs been approved for clinical use. However, they are often associated with a high risk of cytokine release syndrome (CRS) and other adverse events. Since many BsAbs target antigens unique to humans and given the need for new approach methodologies (NAMs) that comply with the 3Rs, it would be valuable to have an in vitro system for predicting immunotoxicity during preclinical development of novel BsAbs. ID.Flow is an ex vivo system using circulating human whole blood at 37 °C, previously evaluated for predicting infusion reactions of monoclonal antibodies and oligonucleotide drugs. The system allows for simultaneous evaluation of multiple immune responses, such as cytokine release, complement activation, activation or depletion of platelets and white blood cells, and coagulation. By using clinically relevant controls, the relative levels of immunotoxicity can be assessed. ID.Flow was used here to evaluate glofitamab, a bispecific T cell engager approved for the treatment of diffuse large B cell lymphoma. Glofitamab is associated with cytokine release syndrome (CRS) including serious or fatal reactions. The efficacy and immunotoxicity of glofitamab were investigated in ID.Flow at concentrations similar to Cmax. As controls, the monoclonal antibodies (mAbs) obinutuzumab and alemtuzumab were used, both of which are associated with CRS in the clinic. Glofitamab resulted in B cell depletion with live B cell counts significantly reduced by 6 hours (60%) and most B cells (90%) depleted by 24 hours. In comparison, B cell depletion using the mAbs occurred with slightly faster kinetics (80-90% depletion by 6 hours). Activation of T cells and B cells was observed in samples treated with glofitamab, but NK cells, granulocytes, and monocytes were also activated. A population of double-positive CD3 + CD22 + events was observed, correlating with the mechanism of action of the drug. Similar to what has been observed in the clinic, inflammatory cytokines (IFN-gamma, IL-2, IL-6, IL-8, and TNF-alpha) were present at very high levels in circulating whole blood treated with glofitamab for 6 hours. IL-6 and IL-8 were induced to similar levels as those seen using alemtuzumab at Cmax, and for IFN-gamma and TNF-alpha, the levels were even higher than those seen with alemtuzumab. IL-2 was, as expected, not produced in response to either obinutuzumab or alemtuzumab, but was induced to high levels (similar to an agonistic anti-CD28 antibody) in the presence of glofitamab. Neither obinutuzumab nor glofitamab had any effect on complement activation in ID.Flow, whereas alemtuzumab induced increased levels of C3a and C5a. In conclusion, ID.Flow can be used to assess the immunotoxicity and potentially the efficacy of bispecific antibodies with glofitamab as a relevant clinical control.